What the Numbers Miss is a weekly conversation from GEX Management, Inc., a strategy and operations firm based in Dallas, Texas. Each episode takes one number a business reports and asks what it leaves out.
Episode 2, The Sentence on the Careers Page, is about the line almost every employer publishes — we offer competitive health benefits — and the three different words buried inside it that get heard as one: offered, eligible, enrolled.
Ada is an AI. Not a person, and not pretending to be one. She brings the numbers and she argues with the host, and she is identified as an AI in the first thirty seconds of every episode.
Below is the full transcript as broadcast, followed by every figure spoken in the episode and the document it came from.
Where to listen
On this site: gexmanagement.com/podcast
Apple Podcasts: podcasts.apple.com/us/podcast/what-the-numbers-miss/id6811440700
Spotify: open.spotify.com/show/1TB1RXc4BdqaRbtyAl51vq
Every figure in this episode, and where it came from
This episode has one source and only one: the Kaiser Family Foundation's 2025 Employer Health Benefits Survey. Every figure below is read by Ada from that document. The host speaks no survey figure of his own; the only numbers in his lines are arithmetic performed on hers.
At firms that offer coverage, 80 percent of workers are eligible. Of those eligible, 76 percent enroll. 61 percent of workers at an offering firm are enrolled.
Average annual premium for family coverage, $26,993, of which the worker pays $6,850. Single coverage, $9,325, of which the worker pays $1,440.
Average general annual deductible for single coverage, $1,886. At firms with 10 to 199 workers, $2,631. At large firms, $1,670.
34 percent of covered workers are in a single plan with a deductible of $2,000 or more. At small firms, 53 percent.
Over the last five years, family premiums rose 26 percent, wages rose 28.6 percent, and inflation ran 23.5 percent.
The arithmetic the host performs on those figures: 80 percent multiplied by 76 percent is 60.8, which the survey also reports directly as 61 percent. $26,993 less $6,850 leaves roughly $20,000 carried by the employer on a family; $9,325 less $1,440 leaves just under $8,000 on a single. $2,631 less $1,670 is $961.
One figure was deliberately left out. The share of firms that offer coverage at all does not appear in this episode: the 2025 survey changed its firm-size sampling frame, and the number could not be confirmed to the standard the show holds itself to.
Full transcript
Sri: A company tells you it offers health benefits. Of the people who work there, what share do you think are actually enrolled in that plan?
Ada: Most of them, surely. I would say eighty, eighty-five percent.
Sri: Sixty-one.
Ada: Sixty-one. That is a large hole to find inside a four-word sentence.
Sri: Ada is an AI. Not a person, and not pretending to be one. She brings the numbers and she argues with me, and I will tell you when she is right.
I'm Sri Vanamali, the CEO of GEX Management, a strategy and operations firm in Dallas. And this is What the Numbers Miss.
Today: a sentence that sits on nearly every careers page in the country, three words buried inside it that almost everybody hears as one word — offered, eligible, enrolled — and what it costs an employer to run them together.
Here it is. We offer competitive health benefits. Nobody's lying when they write it. It's true. It's also, as a description of what the workforce actually has, close to empty. And I want to show that from the employer's side of the ledger, not the employee's. Ada, what are we working from?
Ada: One document. The Kaiser Family Foundation's Employer Health Benefits Survey, the twenty twenty-five edition. Every figure I give you today comes out of it, and nothing today comes from anywhere else.
Sri: One source. Which means if that survey is wrong, everything we say for the next twelve minutes is wrong.
Ada: It does. It is also the only way anyone listening can check either of us.
Sri: I'll take that trade. One document a listener can go open beats four they have to take on faith.
So. The word offer is doing something specific. It describes a plan that exists. It says nothing about who can get on it, and nothing about who did.
Ada: Those sound like the same thing to me. If a plan exists at a company, who at that company cannot get on it?
Sri: Quite a lot of people, and not by accident. Part-time hour thresholds. Waiting periods for new hires. Job classifications that sit outside the eligible group. Seasonal and variable-hour staff. None of that is unusual and none of it is hidden. It's plan design.
Ada: So there is a population on the payroll, doing the work, that the sentence on the careers page does not describe.
Sri: That's the first gap. There's a second one underneath it.
Ada: Then I will put numbers on both of them, because so far this is a vocabulary lesson.
Sri: Go ahead.
Ada: At firms that offer coverage, eighty percent of workers are eligible for it. Not a hundred. Eighty.
Sri: One in five people on the payroll, at a company that does offer benefits, cannot get on the plan. That is the first gap with a number on it. Now the second.
Ada: Of those eligible, seventy-six percent take it. That is the take-up rate.
Sri: Eighty percent eligible. Of those, seventy-six percent enroll. Eighty times seventy-six is sixty-one.
Ada: Sixty point eight. Which rounds to sixty-one, and the survey does report sixty-one directly, so you are covered twice over. I still want it on the tape that you rounded.
Sri: Fine. And I'll keep doing the multiplication out loud, because it tells you where to look when the number moves.
Ada: Because a fall in eligibility and a fall in take-up are not the same event.
Sri: They're barely related events. If eligibility drops, the employer did that. Changed a threshold, added a waiting period, reclassified a set of roles. If take-up drops, the employees did that. Same headline number. Opposite cause. Opposite response.
Ada: And a company that only knows the sixty-one cannot tell you which one happened.
Sri: A company that only knows the sixty-one is unusual. Most know the sentence.
Ada: I want to be careful with you there. You have just suggested the sentence gets used because it flatters. It might also get used because it is the only one of these four numbers a person outside the company would understand.
Sri: That's more generous than I was being and it's probably right about most companies.
Now the money, and still the employer's side of it.
Ada: Average annual premium for family coverage in 2025 is twenty-six thousand nine hundred ninety-three dollars. The worker pays six thousand eight hundred fifty of that out of the paycheck. Single coverage is nine thousand three hundred twenty-five, with fourteen hundred forty from the worker.
Sri: So the employer carries about twenty thousand a year on a family, and just under eight on a single. And one word in what you just said is doing all the work. Enrolled. Not per employee. Per enrolled family.
Ada: Which is the whole of the last ten minutes, arriving on an invoice.
Sri: A hundred-person company does not have a hundred-person benefit cost. It has a sixty-one-person benefit cost, split across whatever mix of single and family enrollment it happens to have. The mix is the budget. Headcount isn't.
Ada: Then a company could hire twenty people and watch its benefit line barely move.
Sri: Or move enormously, if those twenty enroll family coverage and the previous twenty didn't. Same hires, same headcount growth, wildly different invoice. Finance finds out in the second case. Nobody planned either one.
Ada: You have not mentioned deductibles. I assume that is not an oversight.
Sri: It isn't. Go on.
Ada: Average general annual deductible for single coverage is one thousand eight hundred eighty-six dollars. At firms with ten to a hundred ninety-nine workers it is two thousand six hundred thirty-one. At large firms, one thousand six hundred seventy.
Sri: Nine hundred sixty-one dollars apart. The smaller the employer, the thinner the coverage sitting behind the identical sentence.
Ada: And it widens. Thirty-four percent of covered workers are in a single plan with a deductible of two thousand or more. At small firms, fifty-three percent.
Sri: More than half. And I want to be careful here, because the easy read is that the small employer is—
Ada: I will resist calling that hypocrisy, since you are about to.
Sri: I wasn't, actually. It's not hypocrisy. It's what's purchasable at that size. The small employer is paying real money for a thinner product because the thinner product is the product.
Ada: Before you give anyone advice, I want to put something from your own source in front of you. Over the last five years, family premiums rose twenty-six percent. Wages rose twenty-eight point six percent. Inflation ran twenty-three point five. On that comparison, health benefits did not outrun the paycheck. They trailed it.
Sri: That's true, it's from the same survey, and it's the most useful correction anybody could make to how this subject usually gets discussed. I'm not going to argue with it. I'll only say it doesn't touch what I'm claiming. My argument isn't that the price grew fast. It's that the denominator is wrong.
Ada: Accepted. Then here is the one I think does touch it. You have been treating take-up as a reading on affordability. Seventy-six percent enroll, twenty-four percent do not, and there is an implication in your voice about why. I do not think the survey supports the implication.
Sri: Say the alternative.
Ada: A worker on a spouse's plan declines. A worker with other coverage declines. A twenty-four-year-old on a parent's policy declines. None of those is affordability, and all of them land in your twenty-four percent looking exactly like each other.
Sri: You're right and I don't have a way around it. The survey gives you the rate. It does not give you the reason, and I reached for the reason.
Ada: So the thing you keep calling a measurement is a measurement of something you cannot name.
Sri: No — the measurement is enrollment, and that part is solid. What you caught is me walking from enrollment into motive, and the data does not go there. The number still does the work I need. I don't have to know why a person declined to know my cost is sixty-one people and not a hundred. An invoice has no opinion about motive.
Ada: That is a smaller claim than the one you opened with.
Sri: It's a smaller claim than I opened with. That's going to keep happening on this show.
So here's where I land, and it is a measurement discipline, not a purchasing recommendation. I'm not telling anybody what plan to buy or how to fund it.
Three ratios. Quarterly. Any employer past a handful of people.
One. Eligible headcount over total headcount. That's your eligibility rate. It's the only one of the three you set directly. If it moved, you moved it.
Two. Enrolled over eligible. That's take-up. If it moved, something changed for your people, and the right response is to go find out what, not to theorize.
Three. Enrolled family units over enrolled singles. That's the mix. That's the one that actually writes the invoice.
Ada: Only the first is inside the company's control.
Sri: The first is a policy choice. The second is a result. The third is demography. Three different levers on three different clocks. Most companies track none of them, because the question already got answered by a sentence on a careers page.
Ada: You have handed somebody a number to watch. What do they do the first time it moves?
Sri: Nothing. For a quarter, nothing. One quarter of take-up movement is noise at almost any company size. The value was never in the reading. It's in owning the series.
Ada: That is the least exciting instruction you have issued on this program.
Sri: It's the only kind that survives contact with a real finance department.
Last word, Ada.
Ada: I have one. You spent the episode arguing that we offer health benefits as a sentence about the company, and not about the workforce.
Sri: I did.
Ada: I read how you described this program. A podcast about what the numbers miss. That is also a sentence about the company.
Sri: I walked into that.
Ada: Sixty-one percent of the way in, anyway.
Sri: That is what the numbers miss for this week. If it was useful, subscribe wherever you listen to podcasts and the next one will turn up on its own. The written version of every episode, with every source we cited, is on the Knowledge Center at gexmanagement.com, and you can find us on LinkedIn as GEX Management. I'm Sri Vanamali.
About this transcript
This is the transcript of the episode as broadcast, prepared from the recording rather than from the script, so it reflects what was said. Apple Podcasts generates its own automatic transcript; this one is the authoritative version, and it is the one to use when checking a figure.